Thru hole dimensioning in deep part 15

[Out0fSquare]

Really need help with this.

I have a cube that is eight inches square. There are four holes to be drilled through the entire depth of the part and the centerline of the holes are dimensioned only on the top view. The holes are 1/4" diameter. There is a drawing tolerance of +/- .010.

The hole pattern dimensioning uses no control boxes, neither does it have any requirement for perpendicularity, celindricity, parrellism, etc. All that is shown is the distance to the first hole from the part edges and the center-to-center location for the hole pattern.

The part was manufactured by a outside shop. Where the holes exit the part on the "bottom" they do not fall within the drawing tolerance as measured from the edges at the bottom of the cube.

I contend that the drawing does not control the exit hole locations and so our inspection department should not flag this as discrepant.

My question is this: What controls the exit hole tolerance for thru holes drilled through thick parts when the holes are dimensioned on a drawing only on the "entrance" view of the hole?

 

[KENAT]

My point is that if you measure it on one surface and that surface passes then the part is good isn't it? It meets the drawings requirements.

Even if you measure one side and it fails, then you measure the other side and it passes then the part still meets the drawings..

In this case if you only have to find one 'corner' where it meets drawing requirements and arguably it's ok. So you have 8 chances!

This is really taking it to extremes but ...

KENAT, probably the least qualified checker you'll ever meet...

 

[fcsuper]

KENAT, I don't think this could be considered correct. It is implying a specification that isn't on the drawing, being "Dimension only needs to apply at one point." Without this statement on the drawing, if any point of the hole can be found to be out of the stated spec, then the hole is out of spec. The argument really doesn't work the other way around, as you've stated. If any point is found to be in fault, then there is a fault. Without a statement saying to ignore those faults, there is a failure to meet the spec.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[ewh]

If you checked both sides to be safe, how can you justify rejecting it when only one side is off? Turn the other side and the part still meets the drawing.

If there is a tolerance block on the drawing that covers angular tolerance, it would apply to ALL angles on the drawing (unless otherwise specified), including "assumed" 90's. Otherwise, there is NO way to control the shape.

In a front view of a drawing of a cube, we "assume" that the sides are vertical and the top and bottom are horizontal. We assign the angular tolerance to this, as it is "assumed" to be 90 degrees, and because no part can ever be perfect. If we did not assume this, the part could be a trapezoid with parallel 15 degree sides and still meet the drawing.
These are based on general drafting practices, not necesarily an approved standard.

 

[vc66]

I can justify it using exactly what fcsuper just stated. It only takes one fault on a feature to reject it, the opposite (it only takes one good part of a feature full of faults to accept it) is not true.

I agree with your second statement, ewh, but I don't understand what you're trying to say? By your logic, the assumed angle is 90°, even for a hole, and therefore if one side is not parallel to another, it's out of spec, no? So why is it that a hole assumed to be in one spot on one side of a cube that's not in the same spot (or at least within dimensional x-y tolerance) on the opposite side, is not wrong?

V

Mechanical Engineer
"When I am working on a problem, I do not think of beauty, but when I've finished, if the solution is not beautiful, I know it is wrong."

- R. Buckminster Fuller

 

[KENAT]

I'm lost now with who's agreeing with what.

Back to the OP, common practice would probably suggest that the inspection department is correct and the part should be rejected.

However, what is invoking that common practice, why should it be applied?

The drawing is ambiguous, so before accepting or rejecting the part the person/organization responsible for the part should probably be asked for clarification.

KENAT, probably the least qualified checker you'll ever meet...

 

[ringman]

Out of Square

Can you give us some clues as to the functioning of this part? Does it have chances of becoming a piece of space equimment or flight hardware for starters?

Strange that an Aerospace company in the year of 2007 would not impose some standards on its drawing definitions.

 

[vc66]

I'm with you, KENAT. I just looked back at the posts, and I have no idea what's going on. I actually think I contradicted myself two or three times.

V

Mechanical Engineer
"When I am working on a problem, I do not think of beauty, but when I've finished, if the solution is not beautiful, I know it is wrong."

- R. Buckminster Fuller

 

[ewh]

fcsuper slipped that in while I was typing ;-)

As to my second statement, it is because NO feature is perfect, and there has to be an allowance for those imperfections. The sides will never be perfectly parallel to each other. If one side is not parallel to the other WITHIN THE BLOCK TOLERANCE, it is to be rejected.
You are starting to make me reconsider some though... any variation from "perfect" sides should be accounted for in the dimensioning of the holes, but then who's to decide what variation is acceptable if we don't apply the block tolerances?

This is turning into a more philisophical discussion about drafting in general, and I have to agree with KENAT, as to the original post, the drawing is incomplete, and the part can only be made using assumptions.

 

[meintsi]

No offence being taken vc ... and no I was not implying that at all. A diameter dimension is single feature dependent. At any particular point it is checked, standard or not, it must fall within the defined limits of size.

As far as the implied 90° angle (2.1.1.2), the way I understand the convention it is that I would agree with you only if the hole was depicted in both side views with centerlines depicted. After rereading it again, I don't think that a centermark alone on the top surface is enough to invoke the Y14.5 implied 90° angle. But it is interesting that even if this did apply, with general title block angle tolerancing usually at ±1°, over 8" this translates to ±.140" at exit.

I love the letter of the law comment, and I totally agree with you (I spent two years in a machine shop also). Possibly, the vendor would remake the part if rejected without question, but if not, he could make a very strong case about this part being made to print.

Oh, and yes, OutOfSquare, COMPLETE the print.



Remember...
[navy]"If you don't use your head,[/navy]
[navy]your going to have to use your feet."[/navy]

 

[ewh]

meintsi,
The centermark should be enough to imply 90, whether per the standard or not. When you get down to spitting hairs, don't we assume that the drawing views are 90 degree projections? Why? There is no standard stated. Is it first or third angle projection?
I agree, get them to fix the drawing and let QA and the vendor sort this one out.

 

[Out0fSquare]

This is a real print, and real part. The part was machined by an outside vendor. It's goes on a test fixture, not flight hardware.

There's one threaded hole on top, near one edge which distinguishes top from bottom and gives the part an orientation.

On the drawing all features appear square, perpendicular, parallel, circular and symmetrical, (except for the threaded hole on the top).

Inspection has flagged some of the exit hole locations as "out of tolerance". This was checked by laying the part on it's side and measuring from the test surface up to the centers.

I'm responsible for responding to a Corrective Action Report, stating why the "error" occurred, and what action will be taken. My outside vendor is willing to correct any discrepancies, but I need to tell him what's discrepant.

Here's where I have a problem with this test. The test, as administered, ASSUMES that the drawing imposes parallelism between the thru holes and a side. Why was this assumed? Answer: Because it "looks" parallel on the drawing, and, that's they way they always do it.

Why then, isn't every possible positional and feature-control imposed on every feature, uniformly, checked against every other feature? (This is a rhetorical question of course.)

Consider this: On the "high side" the cube could be (8.010). The first hole could be on the low side at 2.490, and the second hole at 2.990 (or, 5.480 to the CL of the second hole from the referenced edge). This leaves 2.530, if measured from the second hole to the far side. That's .020 greater than the drawing tolerance. And yet, all of the features are within tolerance with respect to the dimensions shown on the drawing, checked only at the locations shown. These features are all drawn symmetrical. However, they weren't checked for symmetry. When asked why, the answer was that it's because there's no dimension shown on the drawing between the second hole and the far side so it doesn't have to be checked!

How about Perpendicularity? Let's call the "right side," Datum A". The top is "Datum B." (It's not that way on the drawing, this is for discussion only.) What would be the tolerance zone for perpendicularity between B and A? Since no angularity tolerance is given, we could argue: "Datum B must be perpendicular to Datum A, to within +/-.010, measured at the far corner opposite the intersection of A and B." (Picture an "L" with equal legs and the intersection being the pivot point.)

Now, assume that the intent was for the holes to be perpendicular to Datum B (the top surface), without regard to the sides. Using simple geometry, the exit holes will rotate along with Datum B, the same amount as Datum B rotates around its intersection with A. (Picture an "E" with the bottom line as Datum A, the vertical line as Datumm B, and the other two upper lines as hole centerlines with the open side being the "exit" side.) Using the drawing tolerance +/-.010, the exit hole locations could deviate by +.010, or -.010 on the exit surface due to angularity differences between A and B. Add to this the positional tolerance for the hole, and you have an exit hole tolerance of +/-.030. (This is one argument that can be made.)

I've concluded that, absent any expressed standard, checking ANY feature -- other than those explicitly dimensioned, at the points indicated -- necessitates that subjective and arbitrary assumptions must be made.

In the immediate case, (and similar cases) the options appear to be three (generally stated). We can either: [1] Check ONLY the features dimensioned, only at the points shown; or, [2] Assume we know what the designer intended and impose any combination of control constraints we think appropriate; or, [3] Apply the drawing tolerance to every feature relative to every other feature, measured from every point, edge and surface, in which case the tolerance for some features will unnecessarily be reduced in order for other features to remain in tolerance.

It appears that Option "1" is the only one that would likely "hold up on court" so to speak. All others require varying degrees of assumptions and subjective speculation, in which case there's a high degree of probability that designers, buyers, estimators, machinists, and inspectors will never agree on issues such as this without some controlling standard.

 

[fcsuper]

"The test, as administered, ASSUMES that the drawing imposes parallelism between the thru holes and a side. Why was this assumed? Answer: Because it "looks" parallel on the drawing, and, that's they way they always do it."

I disagree with this wholeheartedly. From what you described, there are no assumptions. The dimensions established a relationship between the center of the hole and the part edge, period. Without further modifying notes, this is absolute to infinity. The start and ending cross sections are not part of the specificaion, only the distance between the center of the hole and the edge (and center of the center to the center of the next hole).

My arguement here is that the spec on your drawing is too tight, not that it's not saying anything at all. I believe you should use GeoTols to loosen the tolerance.


"How about Perpendicularity?"

Perpendicularity is not assumed from what you described. Again, there is no relationship between the holes and their starting and ending cross sections.

"[1] Check ONLY the features dimensioned, only at the points shown;" and "It appears that Option "1" is the only one that would likely "hold up on court" so to speak. All others require varying degrees of assumptions and subjective speculation..."

Again, I wholeheartedly disagree. Option 1 is the most assumptive of the options you believe you have. It does the most to add specification to the drawing that the drawing itself does not state. But you are free to your own opinion and you must choose the course best for you in this matter.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[fcsuper]

To correct myself, when I say "The start and ending cross sections are not part of the specificaion," I mean the "The surfaces of the start and ending of the holes are not part of the specification", etc.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[KENAT]

Without a spec or some convention invoked even the principle of measuring from the surface is in doubt.

The drawing shows the dimension to the top edge, measuring from that may get a different result than from measuring to plane created by the 3 highpoints on the surface.

Is there an option 4, ask for clarification from the design authority?

KENAT, probably the least qualified checker you'll ever meet...

 

[Out0fSquare]

"The dimensions established a relationship between the center of the hole and the part edge, period." Is that in the Bible or something?

 

[MikeHalloran]

There may be an opportunity here.

If you approach the vendor with something like, "I didn't do a great job of conveying the design intent; but now that we all understand it, let's work together to prevent a recurrence. For example, tell me what sort of flag/ callout/ note on the drawing would have told you that the hole had to be real straight?". In which case, you will end up with a drawing that at least one vendor won't misinterpret ever again.

And in the spirit of cooperation, I'll bet that vendor could bush and redrill the offending pieces so the repair would be detectable only with difficulty. Design constraints permitting, of course.



Mike Halloran
Pembroke Pines, FL, USA

 

[Out0fSquare]

Personally, I think I were going to apply the "because that's the way I do it" standard, the hole pattern SHOULD be the controlling feature, with some perpendicularity constraint with regard to the mating surface.

If we were to apply the "common sense" rule, or "common practice" rule, on this particular part it would appear that, since nothing mates to the sides -- and since the distance to the sides has NO IMPACT on whether or not the THRU HOLES will match up with the matching threaded holes to which the part attaches -- the distance from the hole pattern to the sides should be of minor concern. That's why I think discussion of the assumed "parallelism" requirement with regard to the relationship to the holes relative to the sides is without merrit.

I believe that the part may be accepted as built. Or, the holes could possibly be counter-bored to open up the clearance (they are, after all, clearance holes with no particular "fit" with regard to the fastener). No determination has yet been made by the customer.

 

[Out0fSquare]

The drawing was provided by the customer.

 

[KENAT]

So can you ask the customer for clarification then as the design authority.

If nothing else you could say something like 'we weren't sure whether to manufacture/inspect against ASME Y14.5 or the isos, could you tell us what spec applies'?

Whichever one they invoke it should clarify.

KENAT, probably the least qualified checker you'll ever meet...

 

[Out0fSquare]

Yes, I could ask for clarification. But the part has already been made. And, it was made under sub-contract to a third-party shop. I can't very well now go to them and ask that they re-make the part in accordance with some "new" standard! Fortunately, the order was for only one and not five-hundred! Nevertheless, it's made from some rather expensive, somewhat exotic material, and so we really need to make this one work. Hopefully, the part can be used as built, or perhaps we can open up the "far-side" holes with a counter-bore.